Stroke occurs in approximately 11% of children with homozygous sickle cell anemia by 20 years of age. Recently, The Stroke Prevention Trial in Sickle Cell Anemia (STOP) showed that high-risk patients can be identified with transcranial Doppler (TCD) ultrasound and that periodic blood transfusions can reduce the annual incidence of first time stroke in high-risk patients from 10% to <1%. Most children in STOP reverted to normal TCD-signifying change to low stroke risk-while on transfusion. The proposed research (STOP II) builds on STOP to optimize the treatment of high- risk patients. Clinical acceptance of this remarkably effective preventive strategy is limited by lack of information on if and when transfusion can be halted after 30 months of treatment in patients who respond to therapy. One hundred children will be randomized over 36 months within 18 months of follow- up after closing recruitment (i.e., 18-54 months of follow-up per patient). Children with evidence of significant arterial disease on Magnetic Resonance Angiography will be excluded. All patients will receive quarterly TCD examinations. A composite endpoint will be used that will consist of reversion to high-risk (TCD (>200 cm/sec) or clinical stroke. Patients who revert to high risk will be offered return to transfusion. TCD, magnetic resonance studies, key laboratory measures and endpoints will be read and/or adjudicated using centralized blinded procedures proved successful in STOP. Assuming the annual endpoint rate on transfusion remains at <1% as seen in STOP, the study will have 90% power to detect an increase to >10% after halting transfusion. This research will optimize the primary prevention strategy proven effective in STOP with significant potential for children with sickle cell anemia.